Explore the directory

An MCP server that exposes Meta Ads data and management operations to AI assistants, with an installer that can create a Meta access token on your behalf through GoMarble's service.

• MCP servers

A curated list of plugins, skills, MCP servers, patch layers, orchestrators and UIs for DeepSeek Harness, DeepSeek's official open-plugin agent runtime.

• Curated lists

A hands-on repository of runnable tutorials that build an AI agent from first principles — the agent loop, tool executor, memory layer and evaluation harness — before introducing any framework.

• Tutorials

A free, ongoing AI engineering curriculum covering machine learning foundations through LLMs, fine-tuning, RAG, AI agents, inference, evaluation and safety, with a written blog and video for most lessons.

• Tutorials

A transparent proxy, described as Wireshark for MCP, that sits in the real traffic path between an AI client and its MCP servers and shows every JSON-RPC tool call live in the terminal.

• MCP servers
• Debugging

A single MCP server that gives an AI agent eyes and hands inside a running app, driving end-to-end tests by natural language across ten platforms including Flutter, React Native, iOS, Android, Electron and the web.

• MCP servers
• Testing

A local Model Context Protocol server giving AI assistants full CRUD access to QuickBooks Online's 29 entity types and 11 financial reports through OAuth 2.0.

• MCP servers

A Node.js Model Context Protocol server that gives an AI agent access to Mapbox's geocoding, routing, points-of-interest search, isochrones, map rendering and offline spatial calculations.

• MCP servers

A free, local-first desktop app that turns AI coding agents you already run — Claude Code, Codex, Gemini and others — into a coordinated crew of specialists that hand work off to each other on one machine.

• Coding agents

A curated, categorized directory of thousands of APIs for building autonomous AI agents, grouped into Agents, AI Models and MCP Servers, each linking to the provider's own page for pricing and docs.

• Curated lists

A terminal-native, open-source AI coding agent whose requests are routed by a hosted model called Klaatu to one of six cost tiers per message, so routine edits use cheap models and only hard tasks reach frontier-priced ones.

• Coding agents

A shared operating layer for Claude Code and Codex that keeps tasks, project memory, reports and team messages in one place so multiple agent sessions can pick up each other's work.

• Coding agents
• Agent memory

The official Triton Inference Server backend for running large language models on vLLM's engine, handling inflight batching and paged attention for served requests.

• Inference

An MCP server that gives AI assistants structured access to SEC EDGAR filings, financial statements and insider-trading data, with exact numeric precision and links back to the source filing.

• MCP servers

A research codebase for fine-tuning flow-matching robot control policies with online reinforcement learning after an initial imitation-learning stage, evaluated on locomotion and manipulation benchmarks.

• Models
• Papers

An open-source, self-hosted web dashboard for Hermes Agent that adds multi-agent crews, cron-based task scheduling, execution approvals, MCP server management and a visual memory graph.

• Agent frameworks
• Agent memory

A companion library that adds remote skill registries and sandboxed execution of bundled files to Pydantic AI's Agent Skills support, for agents that load instructions and resources only when a task needs them.

• Agent skills

A local runtime, modeled on Ollama, for pulling and serving small decision models that turn a piece of text plus typed questions into calibrated probabilities in milliseconds, through a TypeSafe-compatible API.

• Inference

An academic survey that formalizes the harness surrounding an LLM — execution loop, tool registry, context manager, state store, lifecycle hooks and evaluation interface — as a primary driver of agent reliability, reviewing over 110 papers across 23 systems.

• Curated lists
• Papers

An open-source, self-hosted control plane that coordinates a fleet of AI coding agents — Claude Code, Codex, Gemini CLI and others — from one dashboard, with a shared task board and automatic recovery.

• Coding agents

A curated list of vector database services, libraries, benchmarks, and papers, covering high-dimensional structure and vector search and comparing options across the space.

• Databases
• Curated lists

An OpenAI-compatible API server that runs MLX text, multimodal, image and Whisper models locally on Apple Silicon, exposed through the familiar OpenAI SDK endpoints.

• Inference

A free, open-source desktop app that runs two AI coding agents together — a read-only Mentor that plans and reviews, and an Executor that writes code and runs commands — so each checks the other's work.

• Coding agents

A Claude Code plugin that delegates scaffolding, test generation and other bulk work to the Antigravity CLI running Gemini as a sub-agent, while Claude keeps the judgment calls.

• Agent skills